Case 6
Saving Mothers’ Lives in Sri Lanka
Geographic area: Sri Lanka
Health condition: In the 1950s, the maternal mortality ratio in Sri Lanka was estimated at between 500
and 600 per 100,000 live births.
Global importance of the health condition today: Pregnancy-related complications annually claim the lives
of 585,000 women. Some 99 percent of these deaths take place in developing countries, where women
have a 1 in 8 chance of dying in their lifetime due to pregnancy-related causes, compared with the 1 in
4,800 chance in Western Europe.
Intervention or program: Beginning in the 1950s, the government of Sri Lanka made special efforts to extend health services, including critical elements of maternal health care, through a widespread rural health
network. Sri Lanka’s success in reducing maternal deaths is attributed to broad access to maternal health
care, which is built upon a strong health system that provides free services to the entire population, including in rural areas; the professionalization of midwives; the systematic use of health information to identify
problems and guide decision making; and targeted quality improvements to vulnerable groups.
Cost and cost-effectiveness: Sri Lanka has spent less on health—and achieved far more—than most
countries at similar income levels. In India, for example, the maternal mortality ratio is more than 400 per
100,000 live births, and spending on health constitutes over 5 percent of GNP. In Sri Lanka, the ratio is
less than one quarter of that, and the country spends only 3 percent of GNP on health.
Impact: Sri Lanka has halved maternal deaths (relative to the number of live births) at least every 12
years since 1935. This has meant a decline in the maternal mortality ratio from between 500 and 600
maternal deaths per 100,000 live births in 1950 to 60 per 100,000 today. In Sri Lanka today, skilled practitioners attend to 97 percent of the births, compared with 30 percent in 1940.
T
he reduction in deaths during pregnancy and
delivery has long been held out as a major
international public health goal, but many
countries have had difficulties making progress toward it. Most observers now agree that there are
no quick fixes, and that the solution will come with the
strengthening of now-failing health systems in many
poor countries, building up the training of professional
and paraprofessional health workers, improving access
to both basic and higher-level services, and ensuring the
availability of basic medical supplies and medications to
deal with obstetric problems. The case of Sri Lanka dem-
onstrates how rapidly progress can occur when those
fundamental building blocks are in place.
Mothers Shouldn’t Die in Childbirth
Pregnancy and childbirth are natural events and typically require little or no medical intervention for either
mother or baby. But in about 15 percent of all pregnancies, a severe complication affects the woman—for
example, maternal diabetes or dangerously high blood
pressure sets in, excessive bleeding occurs during childbirth, or the mother suffers from a serious postpartum
Saving Mothers’ Lives in Sri Lanka infection. In about 1 to 2 percent of the cases, women
often require major surgery and may die without effective treatment of these complications.
Over and above the baseline risk of pregnancy, women
are in danger of dying during pregnancy and childbirth
if their general health is poor. Malnutrition, malaria,
immune system deficiency, tuberculosis, and heart
disease all contribute to maternal mortality. In addition,
use of unsafe abortion services is a major risk factor for
maternal death.
Maternal mortality, the death of a woman while pregnant or within about two months after the end of the
pregnancy, echoes through families for many generations. Women who die are in the prime of their lives and
are likely to be leaving behind one or more children—a
loss that places at risk those children’s social development, health, education, and future life chances. The
death of a woman in childbirth is highly correlated
with the survival of the child she is bearing; the risk of
a child dying before age 5 is doubled if the mother dies
in childbirth. At least one fifth of the burden of disease
for children under 5 is associated with poor maternal
health.1 Because poor women are far more likely to die
than better-off women, maternal mortality is one of the
factors contributing to the transmission of poverty from
one generation to the next.
Interventions to detect pregnancy-related health
problems before they become life threatening, and to
manage major complications when they do occur, are
well known and require relatively little in the way of
advanced technology. What is required, however, is a
health system that is organized and accessible—physically, financially, and culturally—so that women deliver
in hygienic circumstances, those who are at particularly
high risk for complications are identified early, and help
is available to respond to emergencies when they occur.
Although some maternal deaths are unavoidable even
under the most favorable circumstances, the vast major-
The official definition of the maternal mortality ratio is the number of
maternal deaths for every 100,000 live births. “Maternal” death refers
to a death during pregnancy or within 42 days after the end of the
pregnancy from a cause related to the pregnancy or its management.
Thus, the death of a pregnant woman from an accident or an infectious disease not specifically related to the pregnancy would not count
in the numerator.
Saving Mothers’ Lives in Sri Lanka
ity can be prevented through systematic and sustained
efforts.
Because of overall high health risks and weak health
systems, almost all maternal deaths take place in developing countries. Ninety-nine percent of the 585,000
maternal deaths each year occur in poor nations.
The extremes tell the story: Women in the poorest subSaharan countries have a 1 in 8 chance of dying during
their lifetime because of pregnancy; Western European
women have a risk of 1 in 4,800. And in the developing world, maternal death is very much the tip of the
iceberg: For each maternal death, somewhere between
30 and 50 other women experience serious injury or infection because of pregnancy and childbirth. In developing countries, more than 40 percent of pregnancies lead
to complications, illness, or permanent disability in the
mother or child.2
During the past several decades, as child health indicators have generally improved in the developing world
and even as fertility rates have fallen, the WHO estimates that maternal mortality has remained relatively
unchanged at a high level. Some countries, however,
have been able to make significant and sustained
progress toward making pregnancy safer for women,
even beyond what would be expected with general
improvements in living conditions and female health.
The lessons from those settings are now informing the
approaches international agencies promote.
Sri Lanka’s Public Health Traditions
Sri Lanka, an ethnically diverse country of almost 20
million people living on a densely populated island in
South Asia, has a storied history of public-sector commitment to human development. Although it is (and
always has been) a poor country, with a current average annual per capita income of $740, the development
of social services even before independence in 1948
has far exceeded the gains made in countries at similar
economic levels. Access to public education was rapidly
expanded during the first half of the 1900s, and schooling of girls has long been much more common in Sri
Lanka than in neighboring countries in the region. As a
result, 89 percent of Sri Lankan adult women are literate,
compared with a South Asian average of 43 percent.3
Health services, too, have benefited from strong publicsector leadership. Going back to the 1930s, the government focused on expanding free health services in rural
areas, with attention given to preventive services and
especially control of major communicable diseases. Financing for this effort was derived largely from income
taxes. Currently, life expectancy in Sri Lanka is 71 years
for men and 76 for women, compared with 57 for men
and 58 for women on average in low-income countries.3
One unusual asset to which Sri Lanka lays claim is a
good civil registration system, which has been in place
since 1867. This system, which first started recording
maternal deaths around 1900, has provided valuable
information for planning and monitoring progress. So,
unlike in most poor countries where maternal mortality
estimates are based on very imperfect sources and methods, Sri Lanka benefits from relatively good data and a
tradition within the public administration of using it.
Elements of Success
Sri Lanka’s success in reducing maternal deaths is attributed to widespread access to maternal health care,
which is built upon a strong health system that provides
free services to the entire population, the professionalization and broad use of midwives, the systematic use
of health information to identify problems and guide
decision making, and targeted quality improvements.
These elements have been introduced in steps, with emphasis first on improving overall (and particularly rural)
access to both lower- and higher-level facilities, then on
reaching particularly vulnerable populations, and later
on quality improvements.
Access
The first challenge in this country that is largely rural
was access. The creation of a basic health service infrastructure, starting in the 1930s, extended access across
rural areas to a range of preventive and curative services, enabling initial improvements in maternal health. At
the lowest level, the infrastructure consisted of health
units staffed by a medical officer, who was responsible
for serving the population within a given area. Within
Our understanding of the pace and causes for decline in maternal
mortality is due to a study by Pathmanathan and colleagues (2003),
which sheds light on the main factors of success. Unless otherwise
noted, information in this case study is drawn from that source.
each of these health areas, public health midwives provided care for all pregnant women.
A viable referral system for both pregnancy-related
and other health problems was established from the
early days. The health units were—and continue to
be—supported by cottage hospitals designed to offer
very basic services; rural hospitals and maternity homes
at a primary level; district hospitals and peripheral units
at the secondary level; and tertiary provincial hospitals
with specialist services, teaching hospitals, and specialist maternity hospitals.
At both the lower and higher levels, the number of facilities was expanded rapidly, increasing from 112 government hospitals in 1930 (about 182 beds per 100,000
people) to 247 hospitals in 1948 (close to 250 beds per
100,000). The secondary and tertiary institutions also
underwent expansion in the 1950s.
No referral system works without accessible transportation—a need that was identified relatively early in
Sri Lanka’s health system development. Between 1948
and 1950, the national ambulance fleet was increased
from 12 to 67 ambulances. All provincial hospitals had
between three and five ambulances each, as did major
district hospitals and those in more remote areas.4
Professionalization of Midwifery
While the basic health infrastructure was being developed, specific attention was paid to the problem of who
would deliver what type of services and, in particular,
how maternal health services would be delivered. The
path chosen was to depend on a large number of clinically qualified midwives. This strategy has proved successful both in Sri Lanka and elsewhere (see Box 6–1).
From early days, public health midwives have underpinned the health unit network. Each midwife serves a
population of 3,000 to 5,000 and lives within the local
area. Midwives’ duties include visiting pregnant women
at home, registering them for care, encouraging them
to attend antenatal clinics, and working with the doctor
who runs those clinics. The midwives are considered to
be one of the most important elements in the excellent
health performance of the country. Supervision and a
referral network back up midwives, who undergo an
18-month training program. They report to supervisors,
Saving Mothers’ Lives in Sri Lanka Box 6–1
The Midwife Approach
The relationship between low maternal mortality and extensive use of professional midwives to deliver antenatal, birthing, and postpartum services, which is seen in the developing world today, has been observed
historically in the industrialized world. In countries where doctors predominantly assisted births in the period around 1920, such as the United States, New Zealand, and Scotland, the maternal mortality ratio was
600 or more per 100,000 live births. During the same period, in countries where doctors and midwives
equally attended births, including France, Ireland, Australia, and England, maternal mortality was lower,
averaging around 500 per 100,000 live births. And strikingly, during the same period, in countries where
midwives attended most births—Norway, Sweden, the Netherlands, and Denmark—the maternal mortality
ratio was very low, between 200 and 300 per 100,000 live births.
Professional midwives have special training to acquire clinical competence, are licensed or registered by
public authorities and are given support, in the form of regular supplies as well as supervision. They also
are linked to a functional referral system, so they know precisely where higher-level care can be obtained
when women face obstetric emergencies.
Midwives are trusted frontline workers who have the distinct advantage of being close to where births are
taking place—within the community—and thus even if they are not called in for each normal birth, they are
available when the unexpected occurs.
Moreover, because midwives can be trained and supported at relatively low cost, and have salaries that
are far lower than medical doctors, the effective use of this cadre of health workers is one of the keys to
saving mothers’ lives within a modest budget.
typically nurse-midwives, who have nursing training in
addition to the basic midwife preparation; the supervisors then report to the medical officer. Established procedures for service delivery and supervision, along with
frequent in-service training, help midwives stay current
and deliver high-quality services.
Importantly, public health midwives are part of both the
health system and their local communities and thus provide a valuable link between the women and the health
units. Even when a midwife does not attend a birth, the
family knows how to find her in the event of a problem.
It is widely maintained that the public health midwives
are key to sustaining the population’s confidence in and
satisfaction with the public maternal health care services.
The growth in the number of midwives was rapid, while
at the same time fertility was falling. As a result, while in
1935 there were 219 live births per government midwife
on average, by 1960 the ratio had fallen to 143 live births
Saving Mothers’ Lives in Sri Lanka
per midwife and by 1995 to 51 live births per midwife
(see Table 6–1).
Largely because of the focus on midwifery—combined
with access to higher-level services—more than in many
other countries, women in Sri Lanka rapidly became accustomed to the notion of attended births and, increasingly, births in hospitals. Up until 1940, skilled attendants assisted only about 30 percent of the births. By
1950, after the implementation of policies to introduce
and expand the cadre of public health midwives, this
percentage had doubled. Concurrently, the proportion
of babies delivered in government health care facilities
increased from 6 percent in 1940 to 33 percent 10 years
later. Currently, skilled practitioners attend to 97 percent
of the births, and the majority are in hospitals.
Information and Organization
Effective management, including the use of information for monitoring and planning, reinforced the two
early building blocks of Sri Lanka’s success in reducing
maternal deaths—access to basic health services and
professional midwifery. In the 1950s, the health education division was formed within the Ministry of Health,
and medical officers of maternal and child health were
designated to coordinate maternal and child health
services in each district.
Quality Improvements, Including Targeting of
Vulnerable Groups
In part because of the information and close monitoring
provided, in the 1960s and 1970s the government identified several ways to improve the system. The Ministry of
Health systematically used maternal death inquiries to
identify problems in the delivery of care—for example,
the reason a problematic delivery was not detected in
time to save a life. The Ministry of Health would then
circulate information about how to prevent similar
problems.
The government’s program to reach women on the tea
estates—farming operations that contracted South Asian
labor in large numbers—provides another example of
a targeted effort to ensure good quality services for all.
Women on the large, privately owned tea estates were
particularly isolated, socially and physically. Once the
estates were nationalized in the 1970s, the government
assumed responsibility for health services, and medical
officers (with transport) and public health nurses established a network of polyclinics to provide integrated
maternal and child health services, including family
planning services, to the tea estates. Estate management gave the women paid leave to attend the monthly
polyclinics.
Bringing these women into the public health system
paid off. Between 1986 and 1997, the number of women
from the estates delivering in hospitals increased dramatically, from 20 percent to 63 percent.
Steady, Impressive Declines
While the maternal mortality ratio (the number of
deaths during pregnancy or immediately afterward
Table 6–1
Development of Government-Employed Birth Attendants,
Sri Lanka, 1930–1995
Year
Live Births per
Government Midwife
Population per 1,000
Government Doctors
Government Nurses per
Government Doctor
Specialist Obstetricians in
Government Hospitals per
100,000 Live Births
1930
405
15.4
n.a.
n.a.
1935
219
n.a.
n.a.
n.a.
1940
n.a.
14.8
n.a.
n.a.
1945
186
n.a.
n.a.
n.a.
1950
163
11.4
1.7
n.a.
1955
157
9.2
2.3
n.a.
1960
143
8.4
2.8
n.a.
1965
n.a.
7.5
2.4
n.a.
1970
n.a.
6.5
2.9
n.a.
1975
n.a.
6.4
2.7
n.a.
1980
125
7.2
3.3
14.0
1985
85
7.4
3.8
15.0
1990
68
7.0
2.7
20.0
1995
51
4.0
2.9
23.0
Note: n.a. = not available
Source: Pathmanathan, Liljestrand, Martins, et al. (2003).
Saving Mothers’ Lives in Sri Lanka divided by the number of births) has persisted at high
levels in many poor countries, Sri Lanka has been able
to halve the maternal deaths (relative to the number of
live births) every six to 12 years since 1935.
In the 1930s, the maternal mortality ratio in Sri Lanka
was estimated to be over 2,000 per 100,000 live births.
By the 1950s, the rate had declined to less than 500 per
100,000. Although data limitations prevent a full explanation of the source of these improvements, it is widely
believed that successful efforts to combat malaria and
the introduction of modern medical practices deserve
much of the credit during this phase.
The steep decline in the maternal mortality ratio that
was observed from the 1930s to the early 1950s has been
attributed largely to the all-out war on malaria.5 DDT
spraying commenced in 1945 and led to a rapid decline
in malaria incidence within a few years. In addition to
the highly successful malaria control program, control
of hookworm infection and general improvements in
sanitation might also have contributed to improvements
in maternal health before 1950.6 Moreover, the rapid decline in maternal mortality during the early 1950s could
be attributed to the introduction of modern medical
treatment, such as antibiotics, through a health service
network established in the pre-1950s era and having
considerable reach in rural areas.
The maternal mortality ratio was halved again during
the following 13 years, up until 1963, when the government made special efforts to extend health services,
including critical elements of maternal health care,
through a widespread rural health network. In the
decades that followed, the public sector systematically
applied stepwise strategies to improve organizational
and clinical management, reducing the maternal mortality ratio by 50 percent every 6 to 12 years. And among
women working on tea estates, the maternal mortality
ratio declined from 120 in 1985 to 90 in 1997 as the
polyclinic system was developed.
In total, this has meant a decline in the maternal mortality ratio from between 500 and 600 maternal deaths per
100,000 live births in 1950 to 60 per 100,000 today.7
Saving Mothers’ Lives in Sri Lanka
Did Targeted Efforts Make the
Difference?
The declines in maternal mortality are clear, as is information about efforts the government made to build
the overall health system and to address the problem
of maternal deaths in particular. A reasonable question to ask, then, is whether the system changes caused
the health improvements, or just happened at the same
time. Tackling this question—using data that span some
60 years—requires piecing together several types of
epidemiologic evidence. And doing so yields a convincing answer.
One way to answer the question of whether system
changes caused declines in maternal deaths is to compare the overall decline in female deaths with deaths due
to maternal causes. Such a comparison is enlightening
because overall female mortality can be assumed to be
related, in large measure, to improvements in living
conditions and in the general health system. In 1950,
maternal deaths accounted for 19 percent of deaths
among women aged 15 to 49 years. By 1996, while both
maternal and all female deaths declined, maternal causes
accounted for only 1.2 percent of all female deaths in the
reproductive age range.
Another way to understand the cause-effect relationship is to look at the changes in maternal deaths due to
individual causes known to be associated with specific
health care delivery strategies. So, for example, deaths
due to hypertensive disease and sepsis—two causes that
are associated throughout the world with lack of access
to skilled attendance—declined dramatically during the
1940s, when emphasis was being placed on increasing
the availability of midwives and skilled attendants at
birth. In contrast, hemorrhage did not decline significantly during the early years studied (1930–1950), when
the major approaches the government took were the
overall development of an accessible health care system,
the control of malaria, and decreasing the proportion of
home births. But between 1950 and 1970, as the government emphasized blood transfusion services and other
strategies to address the problem, maternal deaths due
to hemorrhage decreased from 113 to 45 per 100,000.
Box 6–2
The Honduran Experience
In Honduras, one of the poorest countries in the Western Hemisphere, the maternal mortality ratio declined by 38 percent between 1990 and 1997, from 182 to 108 maternal deaths per 100,000 live births.
This remarkable achievement, which surprised many observers, was the result of a concerted effort by
government officials and development agencies to address maternal mortality.
Although expanding access to essential health services had been a government priority since the 1980s,
a study of maternal health in the early 1990s that revealed a serious problem of maternal mortality served
as a “rude awakening” to the Ministry of Health, according to Dr. Isabella Danel, US Centers for Disease
Control and Prevention expert on maternal health, now posted at the World Bank. This study stimulated
a new focus on safe motherhood programs and the inclusion of specific maternal health priorities in the
national health policy. Importantly, the government used information about differentials among geographic
areas to target its efforts.
By the mid-1990s, a three-part strategy was well into implementation. The first part of the strategy was
a reorganization of health services, intended to increase access to skilled care for pregnant women. This
included the inauguration of community health clinics, with traditional birth attendants supervised by auxiliary nurses; the construction of maternity waiting homes attached to public hospitals; the establishment of
birthing centers supervised by nurse midwives in rural areas; and the expansion of the basic health center
and hospital infrastructure.
As a result of these efforts, between 1990 and 1997, Honduras’ health infrastructure was expanded by 7
new area hospitals, 13 birthing centers, 36 health centers, 266 rural health centers, and 5 maternity waiting homes.
The second dimension of the strategy was the training of health workers in specific areas: Traditional birth
attendants and public health system staff were trained to recognize high-risk pregnancies and deal with
both routine births and obstetric emergencies. Traditional birth attendants were encouraged to accompany women with emergencies to the hospital. The final part of the strategy was community participation,
in which local communities were provided with the opportunity to describe and identify solutions to their
own health problems and, through newly implemented decentralization policies, were given more decisionmaking authority. Although this was very much a government strategy, resources from a variety of donors
were channeled into its support.
Between 1990 and 1997, maternal mortality across Honduras declined, with the biggest reductions in
some of the poorest and most remote areas. So while overall skilled birth attendance changed little during
the period, the number of maternal deaths declined from 381 in 1990 to 258 in 1997 due to better referral of women with complications before, during, and after delivery.
The experience of Honduras, like that of Sri Lanka, challenges the notion that little can be done to act
on the problem of poor maternal health in poor countries. Success depends neither on major technological innovation nor on high levels of spending, but rather on a combination of three factors: government
commitment, which is often spurred by quantifying the problem; targeted actions to improve referrals and
emergency services in hospitals; and expanded access to well-trained birth attendants within the community, supported by higher levels of care.10
Saving Mothers’ Lives in Sri Lanka The main conclusion from this type of analysis is that
the actions of the health system, rather than improvements in general living conditions, led to a large share of
the improvements in maternal health that occurred over
60 years in Sri Lanka. The finding is reinforced by a parallel analysis of a similar trajectory of maternal mortality
decline in Malaysia, which was similarly successful in
achieving a sustained, long-term reduction in maternal
deaths over several decades (although starting from
a lower initial level). In the case of Malaysia,8 public
health researchers have drawn the link between overall
and cause-specific changes in maternal mortality and
implementation of a similar set of strategies: professionalizing midwifery, expanding access, mobilizing women
and communities, and improving management and the
ability to reach the poorest.
In Sri Lanka, the story of improvement in maternal
health goes far beyond the health system itself. Mothers’ health significantly benefited from effective public
investment in basic health services, in improving basic
living standards, and in high levels of female education.
But the Sri Lankan case, like a very different experience
in Honduras (see Box 6–2), reveals ways in which specific strategies to address the problem of maternal deaths
greatly augmented the health benefits that would have
resulted solely from broad improvements in welfare.
Relatively Low Cost
Sri Lanka has achieved much better health status and
steeper declines in maternal mortality than countries at
comparable income and economic growth levels—and it
has done so while spending relatively little on health services, compared with those same countries. In India, for
example, the maternal mortality ratio is more than 400
per 100,000 live births, and spending on health constitutes over 5 percent of GNP. In Sri Lanka, the maternal
mortality ratio is less than one quarter of that, and the
country spends only 3 percent of GNP on health.
Major expenditures, aside from the health infrastructure
that served a variety of purposes other than maternal
care, are on skilled labor. In Sri Lanka, as in other poor
countries, labor is relatively cheap, and in fact, salaries
for civil servants have been declining in relative terms.
The country could afford widespread access to maternal
Saving Mothers’ Lives in Sri Lanka
health care by using a mix of health personnel: Most of
the maternal health workers are well-trained but lowcost midwives and who are described as extremely dedicated. They are closely supervised by nurse-midwives,
who in turn are supported by a small number of medical
doctors.
Most remarkably, Sri Lanka’s success has been achieved
on a decreasing budget. Between 1950 and 1999, the
proportion of the national budget spent on maternal
health services has steadily fallen, from 0.28 percent of
GDP in the 1950s to 0.16 percent in the 1990s. Originally, this was due to efficiency gains made in the 1950s and
1960s. More recently, because salaries of government
health staff have been falling, overall expenditures have
declined. In addition, expenditures on private services
have become relatively more important in the health
sector as a whole: In 1953, an estimated 38 percent of
total expenditures were private; by 1996, about half of
the total spending was from private sources.9
Major Lessons
Sri Lanka’s achievements in reducing the toll of pregnancy have been impressive and correspond to a setting
where the public sector has for many decades placed priority on the population’s health and education. But others can take inspiration from the country’s record: In the
late 1950s, when the first efforts were made to address
the problem of maternal deaths, the GNP of Sri Lanka
was equivalent, in constant dollars, to the national
income of Bangladesh, Uganda, and Mali today and far
lower than that of Pakistan, Egypt, or the Philippines. In
relative terms, Sri Lanka has spent less on health—and
achieved far more—than any of these countries.
The gains that Sri Lanka made were reinforced in many
ways by good education, an emphasis on gender equity,
and broad health system development—but the specific
actions that were taken to solve the problem of maternal
deaths had a separate and identifiable positive impact. In
Sri Lanka, the basic health system served as an essential
platform from which to work but did not itself generate
the impressive results. Those were due to a step-by-step
strategy to provide broad access to specific clinical services, to encourage utilization of those services, and to
systematically improve quality.
References
1.
2.
World Bank. Safe Motherhood and the World Bank:
Lessons from 10 Years of Experience. Washington,
DC: World Bank; 1999.
World Health Organization. Maternal Mortality in
1995: Estimates Developed by WHO, UNICEF, and
UNFPA. Geneva, Switzerland: World Health Organization; 2001.
3.
World Bank. World Development Indicators. Washington, DC: World Bank; 2003.
4.
Wickramasinghe WG. Administration Report of the
Director of Medical and Sanitary Services for 1951.
Colombo, Sri Lanka: Ceylon Government Press;
1952.
5.
Abeyesundere ANA. Recent Trends in Malaria Morbidity and Mortality in Sri Lanka: Population Problems of Sri Lanka. Sri Lanka: Demographic Training
and Research Unit, University of Colombo; 1976.
6.
Wickramasuriya GAW. Maternal mortality and
morbidity in Ceylon. J Ceylon Branch British Med
Assoc. 1939;36(2):79–106.
7.
United Nations. Human Development Report. New
York, NY: United Nations; 2003.
8.
Pathmanathan I, Liljestrand J, Martins JM, et al. Investing in Maternal Health: Learnings from Malaysia
and Sri Lanka. Washington, DC: World Bank; 2003.
9.
Hsiao W. A Preliminary Assessment of Sri Lanka’s
Health Sector and Steps Forward. Cambridge, Mass:
Harvard University and Institute of Policy Studies;
2000.
10. Danel I. Maternal Mortality Reduction, Honduras,
1990–1997: A Case Study. Washington, DC: World
Bank, Latin America and Caribbean Region; 1999.
Saving Mothers’ Lives in Sri Lanka